#Quick Start

This guide will help you launch AIO Proxy within minutes, connect your first model provider, and make a request through the unified endpoint.

If you haven't installed AIO Proxy yet, please refer to the Installation Guide.

#1. Start the Service

Run the following command in your terminal to start AIO Proxy:

aio-proxy run --open

On first launch, it will automatically generate the default configuration file ~/.aio-proxy/config.jsonc in your home directory and open the Dashboard in your default browser:

• API Service Endpoint: http://127.0.0.1:9317
• Dashboard UI: http://127.0.0.1:9317/dashboard

The initial configuration contains no providers. Choose either of the following methods to add your first provider.

#2. Add a Model Provider

#Option 1: Add via Dashboard (Recommended)

1. In the Dashboard, click "Providers" in the left navigation sidebar.
2. Click "Add Provider" in the top-right corner and select API Provider.
3. Fill in the Provider ID, upstream Base URL (e.g. OpenAI or any compatible platform), API key, and exposed model names (such as gpt-5).
4. Save and click the "Test" probe button on the provider card to verify upstream connectivity.

#Option 2: Edit Configuration File Directly

First set your upstream API key in your terminal:

export OPENAI_API_KEY='your-api-key'

Open the configuration file using the built-in command:

aio-proxy config edit

Add your provider configuration inside the providers object:

{
  "$schema": "https://unpkg.com/@aio-proxy/types/config.schema.json",
  "providers": {
    "openai": {
      "kind": "api",
      "protocol": "openai-response",
      "baseURL": "https://api.openai.com/v1",
      "apiKey": "{{env.OPENAI_API_KEY}}",
      "models": ["gpt-5"],
    },
  },
}

Validate and hot-reload the configuration:

aio-proxy config validate
aio-proxy reload

openai is the stable Provider ID; gpt-5 is the model name exposed to clients. You can later add multiple providers for the same model name to enable automatic failover and load balancing.

#3. Send Your First Request

Open a new terminal window and use curl to send a test request:

curl http://127.0.0.1:9317/v1/responses \
  -H 'Content-Type: application/json' \
  -d '{"model":"gpt-5","input":"Introduce AIO Proxy in one sentence."}'

Upon success, the response is returned from your configured upstream provider.

Return to the Dashboard's "Traces" page to inspect full lifecycle details in real time, including Time to First Token (TTFT), token consumption, and round-trip latency.

#Next Steps

• Long-running Daemon: See Background Service or Docker Deployment to run AIO Proxy as an autostart daemon.
• Connect Clients: See Clients & Agent Integrations to point Cursor, VS Code, Claude Code, or business SDKs to the local proxy.
• Explore More Providers: Learn about OAuth Plugin Providers for ChatGPT / Claude subscription accounts, or Multi-Protocol Endpoints.